Vintage Review: NSU Ro80 – The World’s First Rotary Engine Sedan – ‘Wonder-car with dark sides’

Cover of AMS 3/1968 (3 February 1968), showing a red NSU Ro80 on a test track with the headline "Großer Test NSU Ro 80: Wunderauto mit Schattenseiten"

Back in 1968, the most popular auto magazine in Germany presented a comprehensive early test drive of a revolutionary new “Wunderauto”: the NSU Ro80. With its futuristic styling and two-rotor Wankel engine, the Ro80 was a bold move for the Neckarsulm-based automaker, and it seemed poised to make a Tesla-like splash in the European executive car market. auto, motor und sport (AMS) gave the Ro80 a complete shakedown, which we are pleased to present on Curbside Classic in English!

Paul has already given a rundown of the Ro80 story (see Automotive History Capsule: NSU Ro80–The Tragic Automotive Goddess), with Tatra87 offering his take in 2017, and I have covered its history at length elsewhere, so I won’t recap that here, but suffice to say this car’s arrival in 1967 made a big impact. Barring a few minor stylistic eccentricities courtesy designer Claus Luthe, the aerodynamic shape, airy greenhouse, and thoroughly modern engineering — front-wheel drive, fully independent suspension with struts all around, four-wheel disc brakes (inboard at the front) with a load-sensing proportioning valve and dual hydraulic circuits — would have been perfectly creditable even 20 years after its debut.

Press photo showing front 3q view of a metallic blue 1968 NSU Ro 80

© Audi AG Archive

And then of course there was the powertrain! No mere piston engine, but the latest two-rotor KKM612 rotary engine. Despite having a combined chamber displacement of only 995 cc (60.7 cu. in.), it boasted 115 PS (113 net hp, to you and me, 130 hp SAE gross) and 115 lb-ft (155 N-m) of torque, which would have been a respectable output for a 2-liter or 2.5-liter reciprocating engine in those days. This marvel was linked to a three-speed Saxomat semiautomatic transmission with a torque converter and an automatic clutch that disengaged as soon as you touched the shift lever.

All very fancy, but how well did it work? This was the question AMS set out to answer.

A few words of introduction may be in order. auto motor und sport (typically styled in lower case) was established through the merger of two earlier West German magazines, Das Auto and Motor und Sport. The new mag, published every other week from 1951 on, was initially entitled Das Auto, Motor und Sport, but, like the British Autocar and Motor, it dropped the definite article in the early ’60s. It was, and so far as I know remains, the most popular German car magazine.

The Ro80 review, like many AMS road tests of the period, was written by editor Reinhard Seiffert. Seiffert’s reviews remind me a bit of the ones by Road & Track and Car Life publisher John R. Bond from this same period, or of the later work of longtime Car and Driver editors Patrick Bedard and Csaba Csere: sober, astute about technical points, often critical, but not prone to the indulgences or Gonzo Journalism flourishes of a David E. Davis or a Brock Yates (or P.J. O’Rourke, heaven forfend). Seiffert’s Ro80 review is characteristic in style, although this test is about twice as long as the typical AMS road test — clearly, AMS felt this all-new car deserved a thorough airing.

I should add a caveat here that my ability to read German does not even rise to the level of “rudimentary,” so this text is mostly machine-translated, and then corrected by me. There were a few spots where I drew a blank on what Seiffert was attempting to suggest — idiomatic translation is hard! — but I think the results are reasonably coherent; you can judge that for yourself below. Any errors in translation are mine, and are not the responsibility of either the original author or AMS. As we go along, I will also add some editorial commentary.

AMS 3/1968 page 20, p. 1 of NSU Ro80 road test, with the headline "NSU Ro 80 Wunderauto mit Schattenseiten"

The title of the review, “Wunderauto mit Schattenseiten,” means “Wonder-car with Dark Sides.” (Paul, who of course is a native German speaker, suggests that “Wunderauto” might be better translated as “Miracle-car,” but either way, you get the idea.) You might assume this was a reference to the apex seal and eccentric shaft bearing problems that became the early KKM612 engine’s bête noire, but those had not yet emerged when this issue went to press in early 1968. (This road test appeared in AMS 3/1968, with a cover date of 3 February.) Seiffert’s focus is on evaluating how the Ro80 performed and how it measured up to its competition.

The main text begins:

It is a rare occurrence for a completely new car to come onto the market. It is even rarer for a car manufacturer to enter a new price range with such a car. However, it must be considered a real special case when all this is combined with the introduction of a new type of drivetrain.

The Ro80 combines all of these exceptions: it is a completely new concept in which not a single screw has been taken from an existing model. For NSU, it represents a step into the sophisticated “upper middle class,” and is the first series-produced sedan with a Wankel engine.

It is not often that so much courage is invested in a new car. Even in the Citroen DS 19, the engine came from the previous model, as revolutionary as the car was. NSU could not rely on what was already there when converting from a motorcycle factory to a car factory. This was not only a difficulty, but also an opportunity. It was seized — with courage and technical imagination.

Interesting technical details on a car are always an incentive for advance praise. It is no wonder that the Ro80 was voted “Car of the Year.” Everyone is interested in it and so it already has the reputation of a wonder car — as the DS 19 once did. Such a reputation is difficult to live up to. Because in everyday automotive life, it is not only about imaginative construction, but also on sober things like reliability and economy. Here, the miracle cars have to compete with cars that are less innovative but have already passed their practical test. If you buy a car just because you are excited about its shape or technology, disillusionment can follow quickly. So, you have to apply the same practical standards to a “wonder car” as to any other.

Bodywork: Concessions to Streamlining

NSU likes to point out that the body of the Ro80 was not designed by “artistic artists” but was designed according to technical aspects. In fact, it lacks the usual “styling” with its respective fashion trends such as rear trim panels or flared fenders. Nevertheless, the Ro80 body is just as much a product of design as any other. Only here, NSU did not go for the pontoon shape — as with the smaller models — but for the streamlined shape.

A streamlined shape is more technical, more functional because it reduces air resistance. The pontoon shape allows for better use of the car’s floor space and better visibility. Both have their place, and the streamlined shape can even be more attractive to the eye. Streamlined cars look “fast” and are more reminiscent of racing cars than staid pontoon limousines. But they require practical concessions — as does the Ro80.

The Wankel engine offers good conditions for a streamlined shape due to its low height. While 30 years ago the pointed rear end was a feature of the streamline, today, it is recognized that a clean flow at the front of the car is the most important prerequisite for low air resistance in a car. The flow breaks off at the rear anyway, and short, pointed rear ends — such as those on the VW — are pointless. This is why modern racing cars have flat, carefully covered front ends and vertically sloping tails. A similar shape could be achieved on the Ro80 (despite the existing engine), the drag coefficient (Cd) is very low at 0.355.

For 1967–1968, a Cd of 0.355 was outstandingly good. Without getting too much into the weeds on different Porsche variants, I think this was a couple of tenths better than an early Porsche 911, and, like many aspects of the Ro80, it would have been a decent figure in the ’80s.

The flat front section also has practical disadvantages: the headlights are low and get dirty very easily. Even after an initial manufacturing defect in the Bosch headlights (fogging from the inside due to moisture penetrating them) has been eliminated, their light is not one of the Ro80’s best features. The fog lights under the bumper get even dirtier, but the low arrangement is undoubtedly an advantage for fog lights — with the Ro80 you just have to get used to cleaning the headlight lenses frequently.

A phenomenon already known from other aerodynamic bodies is the dirtying of the side windows. In the Ro80, a short drive on a dirty winter road is enough to make the side windows almost opaque. The strong inclination of the front and rear windows causes problems: snow and raindrops tend to stick to the glass and impair visibility, and the standard heating wires in the upper part of the rear window do not do much to change this. Experience has shown that in summer, sunlight can become very annoying if the windows are tilted too far.

AMS 3/1968 page 21, p. 2 of NSU Ro80 road test, including Advantages and Disadvantages table

(In the interests of readability, I’m not going to try to make the paragraph breaks line up exactly with the page breaks, so if you read German and note that the text flow doesn’t quite match up, my apologies.)

The caption for the top photo reads, “The Ro80 is equipped as standard with a tachometer whose ‘red zone’ begins at 6500 rpm. This limit was set with consideration for the service life of the rotor seals.” The inset text box reads:

Advantages

  • Engine is quiet and vibration-free in all speed ranges
  • Good driving performance
  • Safe driving characteristics
  • Very good handling
  • Very good brakes

Disadvantages

  • High fuel consumption
  • Annoying noises from [engine] accessory units
  • Details of body and equipment that need improvement
  • Tight entry front and rear

The main text continues:

The sloping roof pillars also have a negative effect on getting in: the doors are narrow, and the upper part of the entry openings is particularly narrow. In the front seats, you have to get used to the fact that the front roof pillars barely clear your head. In the back seats, headroom is reduced by the sloping roof. These are all concessions to the shape that you have to accept. They could only be avoided if the shape were changed.

NSU will probably not do that, because the shape is one of the features of the Ro80 that caused a sensation. Like Citroen with the DS 19, NSU has managed to find a distinctive body shape that is hardly in danger of becoming outdated. And, although both were based on the idea of streamlining, NSU has managed to do some things better: the beltline is lower, the interior is wider, the trunk is larger. From the front, the car looks extremely low and sporty; the rear is generally not as well done, it looks more massive than the front. Typical of the Ro80 is the side line that rises from the front to the back, which was made possible by the flat front section.

The view to the front is excellent, thanks to the sloping hood and the windscreen, which extends into the roof. Unfortunately, the wipers do not cover the upper part of the windscreen, so the viewing angle is severely restricted when it rains. The roof pillars protruding inwards are a little annoying at the sides, and the view to the rear is average at best. The car’s rear boundaries are not visible, but the front ones are almost visible.

The potential problem with having an airy greenhouse is that it can feel like … an airy greenhouse, which is great in pleasant weather, perhaps not so much in a rainstorm, or in the scorching summer heat. The complaints about not being able to see the tail from the driver’s seat are very familiar issues in modern cars. As many CC readers can attest, part of what made gigantic sedans livable on American roads was that you could often see all four corners, making it easier to place your four-wheeled dreadnought in traffic or in docking maneuvers; some smaller and more rationally sized cars do not score nearly so well on this point.

NSU attests that the body structure is very rigid, which is also guaranteed by sophisticated profiles in the seemingly light roof structure. The post profiles in the roof create “a grid system of roll bars,” while the strong door sills and the relatively wide door panels protect against side impacts. The front and rear sections can, as is common practice today, absorb impact energy by deformation. A good safety factor is undoubtedly the fact that the fuel tank has been placed in front of the rear wheels, in the space least at risk in the event of a collision.

AMS 3/1968 page 22, p. 3 of NSU Ro80 road test

The main text continues:

We were unable to detect any signs of distortion or other rigidity defects, apart from the front hood opening on its own once, which could also have been caused by the locking mechanism. The doors were quite hard to close, particularly in the first of our two test cars, which had been taken directly from the series that had just started production, without being inspected. The body of this car tended to rumble on bad roads, but the second car did so less. The door locks did not make a particularly good impression. They were difficult to open from the outside and the interior locking mechanism sometimes did not work properly. The front hood support is primitive: a weak metal rod that has to be clamped in two places when not in use, which the petrol station attendants usually fail to do. The tank filler opening is unsuitable for automatic filling hoses: The fuel spits back easily, so that refueling can only be done slowly.

This car was a huge step up for NSU in size and price, so while these minor glitches were probably to be expected, they were perhaps a bit hard to swallow given the price. At over 14,000 DM, a new Ro80 in 1968 cost twice as much as a 1.5-liter Ford Taunus 15M. When the Ro80 was sold in the U.S. a year or so later, it listed for $4,995.

Furnishing: Miracles Take a Little Longer

Since even car manufacturers with decades of experience rarely produce a car in which every detail is good from the start, one has to be lenient with a newcomer like the Ro80. Like any other car, it must first go through a maturing process until every detail has reached the highest level of quality and practicality.

As with many car manufacturers, NSU believes it has found a special ideal recipe for the arrangement of the various switches and levers. The driver therefore needs to study it closely, which is particularly difficult if he frequently switches from one car to another. Windscreen wiper levers on the steering wheel, for example, work in all cases known to date in such a way that pulling the lever activates the windscreen washer. On the NSU, however, this sounds the horn — we often confused drivers ahead by honking when we were cleaning the windshield. To wash the windshield, the lever must be pushed forwards rather than backwards. Even if you know this, accidental honking while wiping or washing still happens quite often.

Until the advent of touchscreens (which I still don’t believe belong in a motor vehicle), modern cars had finally, after years of painful development, reached a certain amount of international consensus on control placement, and so it’s jarring to be reminded how much that DIDN’T used to be the case. (The Ro80 control labeling was usefully improved for 1970.)

The caption at the bottom of the page reads, “The function of the switches and signal lamps is not immediately apparent. To the left of the steering wheel is the switch for the indicators and dimming, to the right (almost hidden) is the lever for the wipers, washers and horn. According to auto motor und sport measurements, the usable capacity of the trunk is 348 liters.”

The buttons on the dashboard are not labeled, so they are easily confused at first. Unfortunately, the pleasant English custom of simply having their meaning engraved into the glass of the warning lights has not been adopted either; the heating controls are clearer and sensibly illuminated. Unfortunately, the heat control lever could only be operated with great effort — a defect that often remains in new car models for years. The heating effect itself was good and was sufficient even in very cold weather. The stepless fan achieves a considerable heating and defrosting performance when fully open. The direct ventilation is also good — not only because of its good effect, but also because it can be easily switched off.

Testers in climates less mild than Germany’s were not so convinced of the adequacy of the Ro80’s ventilation system in hotter weather. With the high windshield, I have to assume the Ro80 would be fairly miserable to drive in the summer in Nevada or Texas without a serious aftermarket window-tinting job.

AMS 3/1968 page 23, p. 4 of NSU Ro80 road test, including a B&W ghost illustration showing its major components in situ

The main text continues:

However, we were concerned about the fact that the Ro80 is very sensitive to the windows fogging up. This is probably due to the plastic roof headliner, which not only looks cheap and unsightly, but also lacks any ability to absorb moisture. Buyers of a 14,000-mark car should not be expected to buy such a headliner replacement. The same applies to the poorly attached handles on the doors.

The seats proved to be sufficiently comfortable in continuous use, but there is no doubt that the opportunity to improve overall comfort through the seat suspension was missed: softness and cushioning are not enough to absorb road bumps as well as the Citroen seats. The double adjustability for length and height is good; however, we were not able to use it because we found the seats to be too high even in the lowest position. The upholstery can be of soft wool cloth or perforated vinyl. We would prefer the cloth upholstery because it adjusts better to body temperature and can absorb moisture.

There is nothing wrong with the readability of the two large main instruments (speedometer and tachometer), and the arrangement of the steering wheel and pedals should also meet average requirements. The handbrake lever is conveniently located between the front seats, but the gear lever is too far away. You have to bend over to engage the park position, reverse gear and second gear. The current arrangement does not prevent the passenger from accidentally touching the gear lever and operating the clutch. There is much to be said for locating the gear lever on the steering wheel in semi-automatic vehicles with touch-controlled clutches.

There is plenty of room in the back seats for two people, even if they are not particularly slim. With an interior width of over 140 cm, the Ro80 can be considered a spacious car, if you discount the uncomfortable entry at the front and back. The seat width and knee room in the back are also reasonable for long journeys for four people, and there is a center armrest. A fifth person can be accommodated with the usual loss of seating comfort.

The floor of the car is covered with a carpet that proved to be pleasantly resistant to moisture and easy to keep clean. The high loading sill of the trunk was a problem, as with almost all cars today. Its spaciousness is sufficient: we measured 348 liters of usable space (BMW 2000 388 liters). The Ro80 is not a baggage car, but the interior and trunk meet the requirements of its price class. After the improvements that are still due, this will also be reflected in the equipment.

A trunk volume of 348 liters is about 12.2 cubic feet, which is not particularly big. That’s about the size of the trunk on a 1990s Honda Civic or E90 Toyota Corolla notchback, which I can say from experience is adequate for routine errands, but requires careful planning if you intend a family vacation with more than two people. Since NSU pitched the Ro80 as a car for long trips (for reasons we will get into presently), being no better than the class average in trunk space was perhaps not ideal.

Now, we get into the meatier parts:

Engine: The Future Has Already Begun

It is no secret that there was a “party” at NSU that was against installing the Wankel engine as standard in a sedan. Such internal differences of opinion are unavoidable when it comes to taking a step into new territory. As is well known, caution is the mother of all safety, but it is not always the mother of success. The Wankel development has earned NSU the reputation of being a particularly progressive and energetic automobile factory. This reputation (and thus the share price) could have suffered if the factory had not committed itself to the new engine principle. Especially since the Japanese licensee Toyo Kogyo had already beaten the Neckarsulm factory to the punch with a two-rotor Wankel car. The Wankel future had just begun, and if you say A, you have to say B.

Perhaps more to the point, license fees from the Wankel engine technology had been doing a lot to keep the lights on in Neckarsulm, and so it would have been very troublesome for NSU to not use the technology in its new flagship. (Would you buy a BMW from a salesman who drives a Buick?)

I confess I’m not sure what “party” Seiffert is referring to here, although it may be NSU chief engineer Ewald Praxl, who had been nervous (for what turned out to be good reason) about rushing the Ro80 into production without more time for testing and development. The NSU management board decided they could not afford to wait, which proved to be an expensive mistake.

Toyo Kogyo was what Mazda used to be called, and the two-rotor Wankel model to which Seiffert refers is the Mazda Cosmo Sport, publicly introduced in October 1964:

Front 3q view of a white 1967 Mazda Cosmo Sport on the grass at Pebble Beach, Florida

The text continues:

That there are still valid counterarguments can be seen from the fact that licensee Daimler-Benz is not yet thinking about installing the three-rotor Wankel engine, which is already well advanced in Untertürkheim, as a standard model. And even NSU will in all probability bring out a new mid-range car this year that has a normal reciprocating piston four-cylinder engine. Apparently, it is still considered too risky to equip a car that is produced in large numbers and is intended to be a big seller with a Wankel engine. NSU is therefore pursuing a “two-pronged approach,” which suggests that the Wankel engine is not heading for sole dominance for the time being. This is probably due to reasons inherent in the Wankel principle: NSU had initially decided on a chamber size of 500 cc. The two-rotor engine was created by doubling the Spider’s single-rotor engine. For a car in the 1.5 to 1.8 liter class, the single-rotor engine is too small and the two-rotor engine is too large. Reciprocating piston engines can be varied in size more easily, so it was the simpler and cheaper way for NSU to further develop the existing four-cylinder engine.

But functional reasons speak against considering the Wankel engine as fully developed and suitable for series production. This was already evident with the Spider engine, whose smooth running and elasticity were not quite as good as corresponding four-cylinder engines. Things look better with the two-rotor engine of the Ro80. But here too there are still unsolved problems.

Given what happened later, the last sentence was the understatement of the year, although again Seiffert is talking here about function and driving deportment, rather than reliability.

AMS 3/1968 page 24,showing diagrams of the rotary engine combustion process, the front and rear MacPherson strut suspension, and a view of the complete powertrain and inboard brakes

The captions for the above diagrams, which appear on the left side of the facing page, read:

Left: The four-stroke process in the Wankel engine. In Figure 1, chamber a begins to suck in, chamber b compresses, in chamber c the combustion gases expand and drive the eccentrically mounted rotary piston. In Figure 2, the exhaust slot is open, while chamber a continues to suck in and chamber b continues to compress. In Figure 3, two ignition sparks simultaneously ignite the mixture in chamber b. This stage corresponds to the “top dead center” in the reciprocating piston engine. In Figure 4, the power stroke is in progress in chamber b, while chamber a is changing from the intake to the compression stroke and chamber c is changing from the exhaust to the intake stroke. The next image would again correspond to Figure 1. The eccentric shaft (not visible in the picture) moved by the rotary piston has completed a full rotation, whereas the rotary piston has only completed a third of a rotation. With one full rotation of the piston, the four-stroke process therefore runs three times. Below: The driven front wheels are suspended on triangular wishbones and MacPherson struts. The rear wheels are guided by semi-trailing arms, the spring strut is not involved in wheel location. A small torsion bar operates the valve of the touch-sensitive brake force distributor. Right at the bottom: The two-rotor Wankel engine forms a complete unit with the torque converter, located in front of the front axle. Behind the converter is the plate clutch, which is automatically activated when the gear lever is touched.

AMS 3/1968 page 25, p. 6 of NSU Ro80 road test

Returning to the main text:

In our experience, they [the unsolved problems mentioned in the previous paragraph] have nothing to do with the reliability of the engine. Even the single-rotor engine of the Wankel Spider, with which we drove 50,000 km (see endurance test, issue 7/1966), was — after initial production defects — robust. We had even less of a feeling that something could break after using the Ro80 for four weeks. The lack of mechanical problems is probably one of the main advantages of the Wankel principle: There are no reciprocating parts, and the bearings are not subjected to excessive stress. Valves, valve trains, connecting rods and crankshafts, which are always a problem with reciprocating piston engines, do not exist here. The eccentric shaft, which is comparable to the crankshaft, is subjected to much more even stress. The most critical point is the sealing elements on the side walls and edges of the rotary piston. Their wear was the main reason why the factory limited the speed to 7000 rpm. Nevertheless, the engine is far less sensitive to over-revving than a reciprocating piston engine. The probability of engine failure is correspondingly lower. As far as running smoothness is concerned, the single-rotor engine was troubled by a rumbling running noise and a tendency to shake at low speeds. When the accelerator was released, it tended to jerk. The Ro80 is spared from such phenomena: the engine runs smoothly and quietly even when idling. Jerking when coasting is suppressed by the torque converter. Since the factory does not build the car with a normal clutch, it can be assumed that the jerking has not yet been eliminated in any other way.

Seiffert was correct about the latter point. Early rotary engines were prone to “snatch” on the overrun because lifting off the throttle resulted in hot exhaust gas being drawn back through the port into the combustion chamber. This was most pronounced with a manual gearbox, since the engine would snatch on each gearshift. Given more development time, NSU might have eventually resolved this, making it possible to offer the Ro80 with a conventional four- or five-speed gearbox, but it never happened.

One of the main advantages of the Wankel engine is undoubtedly the extremely low noise level at all speeds. The engine itself is practically inaudible, even at full throttle. The mechanical noises and vibrations that are familiar from a reciprocating piston engine are absent, and the intake, combustion and exhaust noise are also lower due to the different gas exchange processes. With this almost complete absence of noise and vibration, the Wankel engine sets new standards. At high speeds, even eight- or twelve-cylinder engines cannot achieve such smooth running, which can compare to the character of a turbine.

The rotary’s smoothness and quietness at speed was a not-inconsiderable virtue in this era, though perhaps more in West Germany than elsewhere. German cars of this period tended to be under-geared by American or British standards, even on pricey six-cylinder models. While contemporary German engines could usually stand the strain of flat-out running, when pushing hard on the Autobahn, you knew they were working hard, with high levels of engine noise and often of vibration. The creamy smoothness of the rotary was like stepping out of a WW1 biplane with an air-cooled radial piston engine directly into a modern turbofan-powered business jet — maybe not a big deal for Americans, who were usually restricted to 70 mph (110 km/h) or less and accustomed to tall gearing with lots of sound insulation, but near-miraculous for German drivers.

Unfortunately, there is a disadvantage to this: there are singing and humming noises that come from the fan and alternator. They are particularly unpleasant at medium speeds on country roads, but less so when driving fast on the motorway, where wind noise predominates. What is annoying here is not the — low — absolute volume, but the frequency of the noise. Added to this is the fact that the noise does not depend equally on the speed, but on the engine speed. Due to the strong effect of the torque converter, this changes very frequently within the same speed range. We have not yet detected comparable noises in reciprocating piston engines, even those with torque converters. This obviously presents new problems that can only be eliminated by targeted noise reduction on the auxiliary units.

This is a well-known problem: Eliminate one source of noise, and others become more audible. I have struggled with this recently because the Corsi-Rosenthal filter-box fan I had running in my living room expired due to seized fan bearings, which instantly subtracted a layer of white noise that had been masking other mildly annoying low-volume sounds.

The strong torque converter effect — when accelerating or decelerating, the speed can increase or decrease by more than 1500 rpm — results from the initially gently increasing engine torque. While reciprocating piston engines usually deliver good power even at low speeds, the two-rotor Wankel of the Ro80 only reaches its maximum torque at 4500 rpm. The converter had to be designed accordingly: The “stall speed”, i.e. the speed that the engine reaches at full throttle and the car is held in place with the brakes, is relatively high at 2510 rpm. If you accelerate at a low speed, the engine initially revs up before the converter “grips.” This creates a similar feeling to driving with a slipping clutch. Not only is it unusual, but, as we will discuss later, it also has unwelcome effects on fuel consumption.

This high stall speed is roughly comparable to an older Buick Twin Turbine (née Dynaflow) transmission, where the converter is the main source of torque multiplication. A “tight” converter would have had less slippage, but also far less multiplication except when starting from rest. On the Ro80, it would also probably have been less effective at preventing or masking engine snatch, which was also presumably why there was no converter lockup clutch.

The loose operation of the converter sometimes gives the impression of poor acceleration. However, this impression is only justified if you fail to downshift in time. The “selective automatic” of the Ro80 is a normal three-speed gearbox with an additional hydraulic torque converter and a plate clutch that is activated when the gear lever is touched — in other words, the same design as the Porsche Sportomatic or the semi-automatic installed in the VW Beetle. The converter provides a variable ratio in each gear: for example, you can start off in second gear and stay in that gear in city traffic. You can also start off in third gear — but then the acceleration is very poor. If you start off in first gear, the acceleration is excellent. When shifting gears, the acceleration values are such that the Ro80 is classified as a spirited vehicle (see comparison table). It is sufficient to rev the engine up to around 6500 rpm, where the red zone on the tachometer begins. In emergencies, the gears can be shifted much further; between 7000 and 8000 rpm, the engine power hardly drops. If you don’t do it all the time, the engine doesn’t mind.

As the spec table reveals, the AMS Ro80 test car went from 0 to 100 km/h (about 62 mph) in 12.5 seconds. This was quite good for a European sedan of this era (the comparison table on the following page, below, shows how the NSU compared to other cars in its class), and okay by American standards. Despite what nostalgia may suggest, most U.S. cars of the ’60s were not muscle cars, and the NSU had performance roughly equal to a Chevrolet Impala with the base V-8 and automatic. You weren’t going to out-drag a Hemi Charger, but it was perfectly adequate for most needs.

Another point worth noting here is that Seiffert is not nearly so puzzled by the Saxomat transmission as American and British reviewers of the time (or modern commentators) tended to be. While German carmakers had been very slow to embrace fully automatic transmissions (Ford-Werke had then only recently started offering one on the bigger German Ford cars, and Mercedes-Benz hadn’t introduced its own until 1961), there had been a minor vogue for automatic clutches. In the late fifties and early sixties, a Saxomat automatic clutch was offered for about 300 marks on some fairly ordinary middle-class German cars. Seiffert was also no doubt familiar with the semiautomatic transmission on the Citroen DS. Combining an automatic clutch with a torque converter wasn’t as common in Europe, but, as the examples Seiffert cites make clear, it wasn’t unknown either. Americans would probably have been happier with a conventional automatic — the last such clutch/converter combo seen in a U.S. product was the final iteration of the old Chrysler semiautomatic of the early fifties — and by this time, some German buyers might have felt the same way.

AMS 3/1968 page 26, p. 7 of NSU Ro80 road test

The caption for the graphs in the upper right reads, “Power and torque curves in the two-rotor Wankel engine in the Ro80. The torque increases more flatly than with corresponding reciprocating piston engines.” (“Leistung” is “Power,” “Drehmoment” is “Torque.”)

Continuing the main text:

The Ro80 is particularly fast on the motorway, where it accelerates excellently in the two top gears and can easily maintain high speeds. The background noises mentioned are hardly disturbing here, the impression of unusually quiet driving prevails. Our two cars did not quite reach the 180 km/h [112 mph] specified by the factory, they ran just over 177 km/h [110 mph]. It must be remembered that these were brand new vehicles and the measurements were taken in winter temperatures. However, the Wankel engines do not need to be run in; they should be able to deliver almost full power even when new. NSU simply recommends cautious driving for the first 1000 km. The cold start behavior was unusually good; even at minus 20 degrees Celsius, the engines — with the choke pulled — started immediately. This is where the fact that there are two spark plugs per combustion chamber comes in handy. They are arranged in such a way that one provides a “safe” ignition spark for cold starts and driving at partial load, with the other for full load. The plugs are subject to high stress and should be replaced every 10,000 km. NSU points out that only four plugs are needed — two less than in a comparable six-cylinder engine.

Similar arguments are also made for oil and fuel: The engine does not need to be filled with expensive high-alloy oils because there are no demanding lubrication points such as valve tappets or connecting rod bearings. Commercially available HD oil SAE 20 is sufficient in winter and SAE 30 in summer. Premium fuel is also not necessary: Despite the relatively high compression of 9.0:1, the combustion in the elongated combustion chamber equipped with two spark plugs is so even that no pinging occurs. So, you can drive safely with regular gasoline. Nevertheless, the consumption of “operating materials” is the weakest point of the Ro80. This is less true for the oil consumption: we had to top up 3.07 liters of oil over a distance of 2361 km. This corresponds to a consumption of 1.3 liters per 1000 km. In the manual, NSU states 1.0 to 1.5 liters per 1000 km. A certain compensation is that an oil change is only required every 20,000 km. This long interval is possible because the oil content is constantly renewed when the engine is refilled. Some of the oil is lost as a matter of course: it is mixed into the intake gases by a metering pump to lubricate the sealing strips. The Wankel engine is therefore an oil consumer by design, which is why the oil level must be checked every time you fill up. When driving continuously at high speeds, oil consumption increases, as is the case with reciprocating piston engines. In addition to lubrication, the oil also serves to cool the internals of the rotary piston. The engine is cooled externally using water in the usual way.

I don’t have data at hand to say how this oil consumption compares to other German executive cars of the time, although these sound like ’60s Jaguar-ish figures. Of course, that was the nature of the rotary engine design, not a manufacturing or design flaw per se.

The Spider engine’s fuel consumption was already around 10 percent higher than that of comparable piston engines. NSU put this down to the carburetor not being a perfect fit. The two-rotor engine now has a special carburetor system with two Solex side-draft carburetors. But unfortunately another factor that increases fuel consumption has been added: the torque converter. According to the NSU technicians, the engine alone achieves the same fuel consumption as corresponding piston carburetor engines. Whether this is the case in all speed ranges can only be determined by precise comparative measurements on the test bench. The valveless four-stroke process does not rule out the possibility of fresh gas losses occurring, which — at least in some speed ranges — have an adverse effect on economy.

The unavoidable loss of power in the converter and the frequent use of the converter effect definitely contribute to the fact that the Ro80’s road consumption is relatively high. It is particularly high when driving at frequently changing speeds, for example in city traffic. At first we did not want to believe it when our city consumption values were over 20 liters/100 km. However, the exact measurement confirmed these values: When driving normally and not at all hurriedly, the Ro80 reached 20.5 liters/100 km in city traffic, and only when driving extremely cautiously did it manage to get down to 16.0 liters. As soon as routes with a more flowing course were included, such as suburban expressways, the picture became more favorable. Even fast, long-distance trips on the highway are not as expensive in the Ro80 as in city traffic. Consumption values also ranged between 16 and 19 liters on highways and country roads.

Having spent a lot of time delving into European road tests, I have begun to get feel for fuel consumption figures expressed in liters per 100 km. To convert to miles per U.S. gallon, you divide 235.214 by the value in L/100 km — 10 L/100 km is about 23.5 mpg, so 20 L/100 km is … oh dear. The reported 20.5 L/100 km city traffic consumption works out to be about 11.5 mpg, which would be pretty nominal for a full-size American sedan with a large V-8, but was a singularly distressing rate of consumption for a medium-size European sedan of this era. Dark sides indeed!

Here are some comparative performance figures, although regrettably they don’t include fuel consumption.

For Comparison
NSU Ro80 BMW 2000TI Citroen DS 21 Pallas Ford 20M 2300S Mercedes-Benz 230 Opel Commodore
Displacement, cc 2 x 500* 1,990 2,175 2,293 2,281 2,490
Power, PS @ rpm 115 @ 5500 120 @ 5500 100 @ 5500 108 @ 5100 120 @ 5400 115 @ 5200
Price, DM 14 190 11 748 14 740 10 119*** 13 145 10 241***
Acceleration in seconds
0 to 60 km/h 5.7 5.4 6.4 5.8 5.6 5.0
0 to 80 km/h 8.4 7.0 9.9 9.0 8.5 7.9
0 to 100 km/h 12.5 10.5 15.2 13.2 13.3 11.9
0 to 120 km/h 17.9 14.9 21.5 19.9 18.5 17.2
0 to 140 km/h 26.5 21.6 33.2 29.9 27.2 25.2
0 to 160 km/h 41.5 34.0 42.6 42.0
1 km with standing start 34.1 31.6 35.9 34.4 34.3 32.7
Top speed, km/h 177.0 183.0 174.0 171.0 175.0 176.5

* Two sides, 500 cc chamber volume. ** Factory claim. *** Four-door.

AMS 3/1968 page 27, p. 8 of NSU Ro80 road test, with spec table

The caption under the photo at the upper right reads, “Under the hood of the Ro80 there is not much to see of the engine. On the left side of the engine (next to the battery) are the four spark plugs, which are supplied by two ignition coils (top left in the picture). Next to the distributor is the oil dipstick and on the left edge of the picture is the expansion tank of the radiator.”

You could probably puzzle out most of the specs yourself even if you read no German, but while we’re here:

TECHNICAL DATA AND MEASUREMENTS: NSU RO80

ENGINE

Two-rotor four-stroke rotary engine, chamber volume 497.5 cc, compression ratio 9.0:1, power 115 PS at 5500 rpm, maximum torque 16.2 mkg at 4500 rpm, water cooling with thermostat and pump, cooler capacity 8.5 liters, pressure lubrication with oil filter, engine oil capacity 6.8 liters, change quantity 3.6 liters, 2 Solex flat-flow register carburetors 18/32 HHD, mechanical fuel pump, 83 liter tank in the rear, battery 12 V 66 Ah, three-phase alternator 460 watts.

TRANSMISSION

Electrically-pneumatically operated single-disk dry clutch with hydraulic torque converter upstream, fully synchronized selective automatic with 3 driving ranges, gear ratios: 1st stage 2.056, 2nd stage 1.208, 3rd stage 0.788, reverse gear 2.105, axle drive 4.857, converter range 2.2, gearbox oil content approx. 2 liters.

CHASSIS

Self-supporting body, front independent suspension on wishbones and spring struts, coil springs, stabilizer, rear independent suspension on semi-trailing arms and spring struts, coil spring, front and rear hydraulic telescopic shock absorbers, rack and pinion steering with hydraulic power steering, hydraulic foot brake, dual-circuit system, brake booster, front and rear disc brakes, handbrake mechanically acting on additional brake drums on the rear wheels, rims 5J x 14, belted tires 175 SR 14, air pressure with and without load 2.0/1.7 atm.

DIMENSIONS

Wheelbase 2860 mm, track 1480/1434 mm, external dimensions 4780 X 1760 X 1410 mm, internal width front 1475 mm, rear 1475 mm, internal height above rear edge of seat front 915 mm, rear 890 mm, seat depth front 500 mm, rear 475 mm, knee room rear 170—280 mm (depending on position of front seats), turning circle left 12.0 m, right 11.8 m, 3.5 steering wheel turns from lock to lock.

WEIGHTS

Weight fully fueled 1249 kg (of which front axle 779 kg, rear axle 570 kg, weight distribution 62.4:37.6), permissible total weight 1660 kg, payload 411 kg, passenger index number 4.8, power-to-weight ratio fully fueled 10.9 kg/PS, when loaded with 340 kg (4 people with luggage) 13.8 kg/PS.

DRIVING PERFORMANCE

Top speed . . . 177.0 km/h
Acceleration (to effective speeds, full tank, 2 persons):
0 to 40 km/h: 3.2 s
0 to 60 km/h: 5.7 s
0 to 80 km/h: 8.4 s
0 to 100 km/h: 12.5 s
0 to 120 km/h: 17.9 s
0 to 140 km/h: 26.5 s
0 to 160 km/h: 41.5 5
1 km from standing start: 34.1 s

Speeds in Gears

I. stage: up to 80 km/h
II. stage: up to 135 km/h
III. stage: up to 177 km/h

Speedometer Accuracy

Indicated: actual km/h
40: 33,0
60: 52,5
80: 72,0
100: 91,0
120: 111,0
140: 130,0
160: 150,0
180: 168,5
Odometer deviation less than 1%

INTERIOR NOISE

(Measured according to ISO curve A)
Idling while stationary: 59 dBA
At 50 km/h: 68 dBA
At 80 km/h: 71 dBA
At 100 km/h: 73 dBA
At 120 km/h: 75 dBA
At 140 km/h: 76 dBA
At 160 km/h: 78 dBA

CONSUMPTION

(Regular gasoline)
Autobahn section ca. 110 km/h: 14.7 Ltr/100 km
Autobahn section ca. 150 km/h: 19.5 Ltr/100 km
Country road section ca. 70 km/h: 17.7 Ltr/100 km
Country road section ca. 85 km/h: 19.8 Ltr/100 km
Short trips 16—22 Ltr/100 km
Test consumption 19.6 Ltr/100 km

PRICE

Sedan DM 14 190. —
Swiss:
Sedan sfr 18 600.—
Austria
Sedan öS 125 000.—

MAINTENANCE

Inspection every 10 000 km.
Oil change every 20 000 km.
Manufacturer: NSU Motorenwerke AG.

AMS 3/1968 page 28, p. 9 of NSU Ro80 road test

“Verbrauchsvergleich” is one of the many delightful German compound nouns you find often in these reviews; “Verbrauch” is “consumption,” and by context, “fuel consumption,” while “vergleich” is “comparison.” Put them together and you have:

Fuel Consumption Comparison
Consumption in Liters/100 km NSU Ro80 (regular gasoline) BMW 2000 Automatic (premium gasoline
Short trips 19.6 14.4
Autobahn, average speed ca. 140 km/h 18.2 16.7
Autobahn, average speed ca. 110 km/h 14.7 14.7
Country roads, average speed ca. 70 km/h 17.7 14.2
Median 17.5 15.0

Note that the figures above are for a BMW 2000 Automatic and NOT the 2000TI listed in the comparison table above.

As you’ll see, if you either grasp the L/100 km metric or do some conversions, the Ro80 wasn’t dramatically worse than the 2-liter BMW on the Autobahn. (In U.S. terms, the 140 km/h Autobahn averages work out to 12.9 mpg for the R080 and 14.1 mpg for the BMW.) However, the NSU’s greater thirst in traffic or back roads was fairly disastrous: 12 mpg in short trips versus 16.3 for the BMW. Little wonder, then, that NSU tried to rationalize this by suggesting that buyers would mainly use the Ro80 as a highway car rather than for general commuting duties!

These measurements contradict the more favorable values given in the TV test by Rainer Günzler and in some other reports. However, these figures come from cars from the pre-production series, and we do not know the respective measuring conditions. Our measurements were determined using flow meters and by measuring the volume of measuring tanks, so measuring errors are excluded. The consumption curve determined at constant speeds is roughly the same as that determined by NSU for the same car before the test. According to NSU, it is slightly less favorable than the production average, but the differences are only around 0.2 to 0.5 liters/100 km. The test car can therefore be considered normal.

After determining the consumption curve and the various individual measurements, we also carried out a comparison measurement with a BMW 2000 Automatic. This car is 15 PS weaker and almost 100 kg lighter than the Ro80, but has less favorable air resistance. The measurement showed without a doubt that the BMW is considerably more economical (see table). The differences were particularly striking in city traffic, but the BMW also used over three liters less on country roads. Only at the relatively constant highway speed of around 110 km/h did both cars perform exactly the same — here too, it is the speed changes that increase the consumption of the Ro80. The cause was easy to determine acoustically: the torque converter of the automatic ZF transmission in the BMW converted little and “locked” immediately when accelerating, whereas in the Ro80 accelerating always initially resulted in the engine revving up. The driving was done with different drivers and in different order, so that the conditions for both cars were exactly the same.

NSU would rather compare the consumption of the Ro80 to that of large-displacement six-cylinder engines such as the Mercedes 250 S or the Opel Admiral rather than to economical four-cylinder engines such as the BMW 2000. The differences are then smaller, but these cars are undoubtedly superior to the Ro80 in terms of power and weight. The general disadvantage — and probably the most regrettable disadvantage of the Ro80 — is the sharp increase in consumption at changing speeds. Without changes to the engine-torque converter-gearbox combination, this disadvantage is unlikely to be remedied.

This is an important point that is easily overlooked: The high fuel consumption was not only a reflection the inherent thirst of the rotary, but also due to the very loose torque converter. Of course, if you have any experience with manual-shift RX-7s, you’ve probably noted that fuel economy is not that much better with a manual gearbox. Rotary engines have some inherent limitations in terms of fuel consumption, in particular a large combustion chamber surface area that leads to poorer thermal efficiency (and greater hydrocarbon emissions) than a reciprocating engine; it’s an intrinsic flaw.

Driving Characteristics: Triumph for the Ro80

The first test drives had already shown that the Ro80 offers above-average driving characteristics. The test confirmed this impression — here the Ro80 has its trump cards to play.

The least of these advantages is driving comfort, which is good but not unusually good. After a longer period of familiarity, it becomes clear that the Ro80 can handle larger bumps smoothly thanks to its long wheelbase, but that the suspension lets short shocks through quite clearly. This is particularly noticeable on the rear wheels, which is why the rear seats are less comfortable to drive than the front. Improvements can still be made here, particularly in the seat suspension.

A big plus of the Ro80 is undoubtedly its driving safety. We felt extremely comfortable in the Ro80 in the critical road conditions of the winter of 1967/68. Good straight-line stability, slightly understeering cornering behavior without a tendency to slide away. Insensitivity to crosswinds are the most important characteristics of its driving behavior, although this does not describe it fully. The power steering plays a very important role, making the car as maneuverable as a small car without being the least bit nervous. It provides a feeling of confident and effortless driving that is only available in a few cars. The main benefit of front-wheel drive is that the directional stability is maintained even when accelerating on slippery surfaces and in snow — after a few winter weeks in the Ro80, it is hard to avoid the realization that front-wheel drive offers enormous advantages even in larger cars. Even if you enjoy keeping rear-wheel drive cars balanced on snow and ice, you are happy to forgo this pleasure and enjoy driving with absolute directional stability.

An oversteering reaction when suddenly releasing the accelerator in curves does not occur because the torque converter softens the transition from pulling to pushing. The long wheelbase and the slightly understeering driving behavior also contribute to this: There is no noticeable difference in cornering behavior with and without accelerator or when transitioning from one state to the other. You hardly notice that the car understeers when cornering because the power steering greatly reduces the amount of force needed to steer. Despite this, you still have enough contact with the road — it’s hard to imagine a more trouble-free chassis design.

Rear-drive purists may scoff at Seiffert’s comments about the lack of oversteer, but for a long time, it was not at all uncommon for FWD cars to oversteer enthusiastically on a trailing throttle, especially if they had any modicum of rear roll stiffness. On the better hot hatches of the ’80s, this could be fun, if you were prepared for it, but even stolid FWD sedans were sometimes prone to this behavior, with occasionally alarming results. The early Ford Taunus 12M (P4), née Cardinal — with which Seiffert was very familiar — was notorious for its nasty trailing-throttle behavior, while the revised suspension of the subsequent P6 version overcorrected by dialing in so much understeer you had to wrestle with the wheel to maintain your line through a curve. Well-considered mild understeer with no nasty surprises is something we generally take for granted on decent modern FWD cars (and of course newer cars have stability control), but this was a revelation at the time.

Grip was not vast, of course, with 175 SR14 tires, although larger wheels and tires with more unsprung weight would likely have imposed penalties in ride quality.

AMS 3/1968 page 29, p. 10 of NSU Ro80 road test with test conclusions

Continuing the main text:

The exceptional driving safety is completed by the excellent braking system. Even on slippery roads and in curves, braking is still surprisingly good, as premature locking of a wheel is prevented by the load-dependent brake force distributor. This valve, connected to the rear suspension, reduces the braking force applied to the rear wheels during rebound and increases it during compression. When the trunk is fully loaded, the rear wheels are used more to brake than when the car is empty, and when the weight of the car shifts forwards when braking hard on a road with good grip, the proportion of braking force on the front wheels is increased accordingly. Another advantage of the braking system is the dual-circuit arrangement, with one circuit acting on the front and rear wheels and a second circuit acting only on the front wheels. If one circuit fails, the far more important braking ability of the front wheels is retained. In terms of safety, this design is far superior to the usual division into front and rear brake circuits. Finally, the braking effect of the four disc brakes themselves, the dosage of the brake booster and the effect of the separate drum parking brake are also excellent. There are not many braking systems that leave so little to be desired.

At the risk of reviving an old argument, this is the kind of brake system that the FWD Cadillac Eldorado and Oldsmobile Toronado SHOULD have had from the start.

There is no question that rarely has such a remarkable new car been brought onto the market as the Ro80; but there is also no question that, alongside its convincing advantages, it also has some drawbacks. These include the imperfections of the body and equipment, but especially the high fuel consumption. We are reluctant to pass judgment on the suitability of the Wankel engine for cars. However, there is no question that some development work will still be necessary to reduce the consumption disadvantage to an acceptable level.

There are many reasons to buy this car — in addition to the shape and the excellent driving characteristics, the advantages of the Wankel engine can also be cited in its favor: the absence of vibrations and insensitivity to speed, the safe cold start, and probably also a low susceptibility to malfunctions and defects.

As a buyer, you will definitely have to deal with the consumption disadvantage. It is particularly noticeable if you mainly drive short distances. If you mostly drive long stretches on the motorway, however, you will not notice it as much — the Ro80 is primarily recommended as a fast and comfortable motorway car. Investing 14,000 marks in this is probably less a question of penny-pinching common sense than of the enthusiasm that this bold and unconventional car is able to arouse.

The “low susceptibility to malfunctions and defects” didn’t pan out, of course, and mechanical issues ultimately proved more costly than the fuel consumption, not least for NSU itself.

Finally, we have the comprehensive summary:

RESULTS

  • Bodywork Streamlined, stylistically unique body shape. Visibility is still very good to the front and sides, average to the rear. Entry is impaired by narrow doors and sloping roof pillars. Adequate interior and trunk space, headlights and side windows sensitive to dirt.
  • Equipment Extensive equipment, but not satisfactory in every detail. Heated rear window as standard. Seats with many adjustments but not ideal comfort. Primitive roof liner, poorly attached door handles. Door locks are stiff and do not work satisfactorily. Good heating and ventilation.
  • Operation Switches, levers and warning lights are confusing, center gear lever is too far forward and disorganized. Handbrake lever conveniently located between the seats.
  • Transmission Three-speed gearbox and compliant torque converter require some getting used to. Starting off in second gear and driving without gear shifting in city traffic is possible.
  • Engine Vibration-free, quiet engine running at high speeds, but singing and humming noises from the auxiliary units. Unusual insensitivity to speed. Low torque in the lower range is compensated by the converter.
  • Performance Good acceleration, top speed almost 180 km/h. Good power reserves in all driving conditions, wide gear range possible.
  • Consumption Fuel consumption increases sharply when driving at varying speeds, above-average consumption values, especially in city traffic. Oil consumption between 1 and 2 liters per 1000 km due to design.
  • Driving Characteristics Safe and unproblematic driving behavior, low wind sensitivity, stable, slightly understeering cornering behavior, good winter capability.
  • Driving Comfort Good overall comfort due to long wheelbase and long-travel suspension. Responsiveness to short bumps and seat suspension not entirely satisfactory. Low noise level. Annoying wind noise only in crosswinds.
  • Steering Excellent handling thanks to smooth and direct, but not nervous, power steering. Relatively small turning circle.
  • Brakes Easy to operate and control brakes with consistent and reliable effect. High braking safety thanks to load-dependent brake force distributor and two front brake circuits. Good handbrake.

With the benefit of hindsight, one could of course compare this test to the enthusiastic early road tests of other conceptually impressive but fatally flawed cars (the GM X-bodies, for instance). There’s some truth to that, certainly, but it also points to the basic limitations of the road test format — there is only so much you can determine about a car even in a comprehensive test like this. AMS did of course follow up with a 50,000-km (31,000-mile) long-term test of the Ro80 (issue 14/1969).

However, this test gives a pretty clear view of how critics saw the NSU back when it first appeared, with its shiny futuristic sheen (and complex three-piece apex seals) still intact.

Related Reading

Rotary Revolutionary: The NSU Ro80 by Aaron Severson (on Ate Up With Motor)
Automotive History: German Deadly Sins (The Neckarsulm Chronicles, Part 3) – The NSU Ro 80, Todsünde Durch Technik by Tatra87